基于IFOA-GA任务调度算法在云计算MapReduce模型中的研究

针对传统的云计算任务调度算法存在效率低、利用率不高的问题，采用改进的果蝇算法（improved fruit fly optimization algorithm，IFOA）和遗传算法（genetic algorithm，GA）融合的算法用于处理任务调度。首先，将任务调度转换为DAG（directed acyclic graph，DAG）并通过Kruskal算法将任务调度顺序进行化简；其次，针对果蝇算法的种群采用正交数组和量化技术进行初始化，对果蝇算法边界进行处理，对探索步长进行动态调整，并使用GA算法对个体选择进行选择处理；最后，将融合后生成的算法IFOA-GA用于仿真平台中的云计算任务调度，相对于IGA、IFOA，IPSO算法在QoS的四个指标对比中具有一定的优势，说明IFOA-GA算法能够有效地提高云计算调度效率。

Research on cloud computing MapReduce model based on IFOA-GA task scheduling algorithm

Aiming at the low efficiency and low utilization rate of traditional cloud computing task scheduling algorithms, this paper proposed an improved algorithm using the improved fruit fly optimization algorithm(IFOA) and genetic algorithm(GA) for task scheduling. Firstly, the task scheduling is converted into a DAG(directed acyclic graph), and the task scheduling sequence is simplified through the kruskal algorithm. Secondly, the population of Drosophila algorithm is initialized using orthogonal arrays and quantization techniques. The boundaries of the Drosophila algorithm algorithm are processed, the exploration step size is dynamically adjusted, and the individual selection is processed using the GA algorithm. Finally, the fusion algorithm IFOA-GA is used in cloud computing task scheduling in the simulation platform. Compared with IGA, IFOA and IPSO algorithm, it has certain advantages in the comparison of four indexes of QoS, which shows that the IFOA-GA algorithm can be effective in mproving cloud computing scheduling efficiency.